Haven't read the work, but if I can extrapolate based on assumption, this seems like something that makes sense in an innate way.

Colour would be the best example. And I think it's an interesting one. The utility in recognising district colours is fairly obvious. Our conscious and memory need a way to label the experience of encountering different wavelengths of light, Otherwise you wouldn't be able to recognise them again surely? You at least need a form of language internally to have the ability to recognise a pattern you've experienced. To me that speaks to the utility of internal dialogue/monologue.

Your own experience of a specific colour can differ wildly from another person's. However, because the wavelength is the same, you can attach a common label to it.

The question of which originated first is interesting to me, but because of the further point, a fundamental system of attaching common labels must exist. Kids can often sort objects in categories before language skills develop.

Seems to me that we do have a universal internal language innate to all of us and we learn a common language later. It also stands to reason that the origins of external language must be based on ancestral internal language.

Perhaps those without verbal internal monologue/dialogue have a more persistent innate language, that is not overwritten by common external language?

/Ramble